| Authors | DR. JYOTI. Y. KULKARNI |
|---|---|
| Article Type | Research Article |
| Language | English |
| Journal | North Asian International Research Journal of Multidisciplinary |
| ISSN | 2454-2326 |
| Volume | 10 |
| Issue | 1 |
| Pages | 1-11 |
| Publication Year | 2024 |
| Publication Date | January 01, 2024 |
| DOI URL | https://doiglobal.org/10.2024/NAIRJCMD.032 |
Food-grain adulteration affects quality assurance, consumer confidence and food safety. Conventional inspection based on manual sorting, visual examination and laboratory testing remains important, but it can be slow, subjective or unsuitable for rapid high-volume screening. This paper proposes a computer-vision framework for intelligent detection and classification of visible adulteration in food grains using deep learning. The proposed workflow covers image acquisition, annotation, preprocessing, augmentation, convolutional neural-network (CNN) feature learning, transfer learning, classification and explainability. The framework is designed for grains such as rice, wheat and pulses and for visually observable foreign matter or abnormal kernels, including stones, husk, damaged grains, discoloured grains and selected look-alike adulterants. The paper distinguishes clearly between visual adulteration that can reasonably be detected from RGB images and chemical or microscopic adulteration that requires laboratory or multispectral methods. Areproducible experimental protocol is proposed using stratified train-validation-test partitions, class-wise precision, recall, F1-score, confusion matrices and macro-averaged metrics. MobileNetV3, EfficientNet-B0 and ResNet50 are proposed as comparative transfer-learning baselines, with an ensemble or optimized lightweight model considered for deployment. No fabricated experimental accuracy is reported; instead, the results section defines the tables and statistical evidence that should be populated after actual model training. The study concludes that deep learning can serve as a rapid screening and decision-support layer for grain quality control, but it should complement—not replace—validated food-analysis methods and regulatory laboratory testing.
Total Views
Total Downloads
Publication Year
DOI Status
This scholarly article has been successfully registered and assigned a persistent DOI through DOIGLOBAL.
DOI: 10.2024/NAIRJCMD.032
Journal:
North Asian International Research Journal of Multidisciplinary
DOI Provider:
DOIGLOBAL
This scholarly record is permanently registered through DOIGLOBAL DOI Infrastructure and remains accessible using its persistent DOI identifier.